DrawContext::run_gpu_pass end-to-end demo.
Shows the same pattern a user would reach for if they wanted to
retain their own GPU buffers for instanced rendering of a large
object scene: a CustomRenderPass implementation owns its
wgpu::RenderPipeline, its base-quad wgpu::Buffer, and its
per-instance wgpu::Buffer. All three are constructed once in
initialize and reused on every frame. The pass is wrapped with
blinc_gpu::GpuPass::new(...) so it can be passed through
DrawContext::run_gpu_pass from inside a regular canvas()
closure.
Tip: Some demos are best viewed in a full browser window. Click “Open in a new tab” below for the full experience.